BASE

BASE is a consistency model for distributed and NoSQL systems that relaxes the strict guarantees of ACID in exchange for higher availability and partition tolerance. The acronym is a deliberate contrast to ACID.

Basically Available

The system guarantees availability — every request receives a response (success or failure), even during partial system failures. Data is replicated across multiple nodes so the system remains operational when individual nodes go down.

Soft State

The state of the system may change over time, even without new input. Replicas may not be immediately consistent — data written to one node may not yet be propagated to all replicas.

Eventual Consistency

If no new updates are made, all replicas will eventually converge to the same state. The system does not guarantee when this will happen, but it guarantees it will happen.

sequenceDiagram
    participant Client
    participant Node_A as Node A
    participant Node_B as Node B
    participant Node_C as Node C
    Client->>Node_A: WRITE x = 42
    Node_A-->>Client: ACK (success)
    Note over Node_A,Node_C: Replication in progress...
    Node_A->>Node_B: Replicate x = 42
    Node_A->>Node_C: Replicate x = 42
    Note over Node_A,Node_C: Eventually consistent

When BASE Makes Sense

Scenario	Why BASE
Social media feeds	Slight delay in showing new posts is acceptable
Shopping carts	Cart state can be reconciled later
DNS propagation	Propagation delay is acceptable
Analytics / metrics	Approximate counts are fine in real-time
Content delivery	Stale cache is better than no response

BASE vs ACID

Property	ACID	BASE
Consistency	Strong (immediate)	Eventual
Availability	May sacrifice for consistency	Prioritized
Concurrency model	Locking / MVCC	Optimistic, conflict-free replicated data
Typical systems	PostgreSQL, MySQL, Oracle	Cassandra, DynamoDB, CouchDB
Use case	Financial transactions, inventory	High-scale web apps, IoT

Tunable Consistency

Modern systems blur the line — many NoSQL Databases let you tune consistency per query. See CAP Theorem for the theoretical foundation behind these trade-offs.